This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sankoorikal, B.-J. Articles by Dannies, P. S. Search for Related Content PubMed PubMed Citation Articles by Sankoorikal, B.-J. Articles by Dannies, P. S.

Aggregation of Human Wild-Type and H27A-Prolactin in Cells and in Solution: Roles of Zn²⁺, Cu²⁺, and pH

Department of Pharmacology, Yale University School of Medicine, New Haven, Connecticut 06520

Address all correspondence and requests for reprints to: Dr. Priscilla S. Dannies, Department of Pharmacology, Yale University School of Medicine, 333 Cedar Street, New Haven, Connecticut 06520-8066. E-mail: . priscilla.dannies{at}yale.edu

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Aggregation of hormones is an important step in the formation of secretory granules that results in concentration of hormones. In transfected AtT20 cells, but not COS cells, Lubrol-insoluble aggregates of human prolactin (PRL) accumulated within 30 min after synthesis. Aggregation in AtT20 cells was reduced by incubation with 30 µM chloroquine, which neutralizes intracellular compartments, and was slowed by incubation with diethyldithiocarbamate, which chelates Cu²⁺ and Zn²⁺. H27A-PRL aggregated in AtT20 cells as well as wild-type PRL, indicating that a high affinity Zn²⁺-binding site is not necessary. In solution, purified recombinant human PRL was precipitated by 20 µM Cu²⁺ or Zn²⁺. In solution without polyethylene glycol there was no precipitation with acidic pH alone, precipitation with Zn²⁺ was most effective at neutral pH, and the ratio of Zn²⁺ to PRL was greater than 1 in the precipitate. In solution with polyethylene glycol, precipitation occurred with acidic pH, precipitation with Zn²⁺ occurred effectively at acidic pH, and the ratio of Zn²⁺ to PRL was less than 1. The aggregates obtained in polyethylene glycol are therefore better models for aggregates in cells. Unlike human PRL, aggregation of rat PRL has been shown to occur at neutral pH in cells and in solution, and therefore these two similar proteins form aggregates that are the cores of secretory granules in ways that are not completely identical.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
NEUROENDOCRINE CELLS STORE protein hormones in concentrated forms in secretory granules, so that large amounts of hormones are rapidly available when needed. The proteins in the cores of these granules are concentrated enough that they may be isolated by centrifugation as insoluble aggregates (1, 2). Aggregation may function not only as a means to concentrate proteins, but also make an essential contribution to the mechanisms of forming secretory granules. Such a role is indicated by the morphology of PRL-producing cells in the pituitary glands of lactating rats, in which aggregates of PRL are scattered throughout the trans-Golgi lumen (3). Rambourg and co-workers (3) concluded that the trans-Golgi layer ultimately is consumed by budding of many vesicles that are too small to contain the aggregates of PRL, leaving behind the aggregates, which become the dense cores of granules. Such an interpretation of the morphology suggests that secretory granules containing PRL are left after the removal of everything soluble, so that the formation of PRL aggregates is a primary event in the formation of secretory granules. Understanding the basis for aggregation is therefore important for understanding how secretory granules form in neuroendocrine cells.

PRL is a good candidate for investigating hormone aggregation for two reasons. First, PRL is a simple molecule. PRL is a relatively small protein (22,500 Da) and is active as a monomer, and the bulk of PRL is not proteolytically processed or modified beyond removal of the signal sequence necessary for transport into the endoplasmic reticulum. Although covalent modifications of PRL in the secretory pathway have been reported, these do not always occur and are not necessary for concentration and storage of PRL (4). Second, it is possible to follow the aggregation of PRL in cells based on the acquisition of Lubrol insolubility by hormones with time after synthesis. Rat PRL is soluble immediately after synthesis when GH₄C₁ cell membranes are dissolved by the detergent Lubrol, but becomes insoluble as it forms the aggregates that are the dense cores of granules (5). It is therefore possible to investigate PRL aggregation not only in solution, but also in cells.

Previous work on PRL aggregation in both solution and cells had left several questions unresolved. One is whether the change in pH that occurs in the secretory pathway is necessary for hormone aggregation. The environment of proteins becomes more acidic as they progress through the secretory pathway; the pH drops from neutral in the endoplasmic reticulum to near 6 in the lumen of the trans-Golgi layer and to 5.5 in secretory granules (6, 7, 8, 9, 10). In solution, aggregation at mildly acidic pH, often enhanced by the presence of Ca²⁺, is a property shared by many proteins that are stored in granules (11, 12, 13, 14, 15, 16, 17, 18). In GH₄C₁ cells, however, aggregation of rat PRL is not prevented when acidic intracellular compartments are neutralized, suggesting that other factors are important (5).

A second question is whether there is a role of divalent cations in hormone aggregation. There are high concentrations of Ca²⁺ in the secretory pathway; total Ca²⁺ in the secretory pathway of PC12 cells is at least 10 mM (19), and Ca²⁺ causes some secretory proteins to precipitate in solution (12, 14, 15, 16, 18, 20). Neuroendocrine cells also have Cu²⁺ and Zn²⁺ in their secretory pathway. Zn²⁺ is present in the trans-Golgi region and secretory granules of the rat anterior pituitary gland (21). PRL has a Zn²⁺-binding site with an affinity of about 1 µM (22, 23), and this binding site has been proposed to be important for aggregation (24). Cu²⁺ must also be present in the trans-Golgi region and secretory granules, as the amidation enzyme peptidyl glycine -amidating monooxygenase requires Cu²⁺ for activity (25, 26). In addition, a mammalian homolog of Cox17p, a yeast protein that participates in the transport of Cu²⁺ into mitochondria, is present in higher amounts in the pituitary gland than in other tissues (27). There are low amounts of Cox17p in fibroblasts in the area of the mitochondria, but much higher amounts in AtT20 and GH₄C₁ cells, primarily in the area of the Golgi complex (27).

A third question is whether aggregates in solution resemble what occurs in cells. Proteins may aggregate differently in cells than they do in solution because they are present at very high concentration in cells; the concentration of insulin in granules has been measured directly and is 42 mM (28). Proteins at such high concentrations may not behave as they do in dilute solution, because the thermodynamic activity of proteins for reactions such as self-association dramatically increases in an exponential fashion at high concentrations (29, 30, 31). Such increases in the thermodynamic activity of proteins have also been caused by adding large inert molecules, such as polyethylene glycol (PEG), to solutions, a process known as crowding or confining proteins (29, 30, 31); such additions not only may increase the thermodynamic activity of proteins, but may also influence the interactions that predominate (31). Such effects raised the possibility that aggregates from dilute PRL solutions might not mimic those found in cells. We have investigated the roles of pH and divalent cations on aggregation of human PRL in neuroendocrine cells and in solution in dilute and crowded conditions.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Purification of human PRL
PRL was produced in BL21 DE 3 Escherichia coli bacteria from inclusion bodies as previously described (32, 33) with the following modifications. Bacteria were resuspended in 50 mM Tris (pH 7.5), 20 mM NaCl, 0.1 mM phenylmethylsulfonylfluoride, 10 µM leupeptin, 10 µM pepstatin, 0.2% NaN₃, and 1.25 mg lysozyme (Sigma, St. Louis, MO) per gram wet pellet. The mixture was incubated on ice for 30 min, subjected to three freeze-thaw cycles, and brought to 20 mM MgCl₂. Deoxyribonuclease I (Sigma) was added to give a final concentration of 10 µg/ml, and the mixture was incubated for 30 min. The inclusion bodies were recovered by centrifugation, and unfolding and renaturation were carried out as previously described (32, 33). Urea and mercaptoethanol were removed by dialysis against 50 mM NH₄HC0₃, followed by a final dialysis against 10 mM Tris and 10 mM bis-Tris propane, pH 8. The solution was applied directly to a POROS diethylaminoethyl column and eluted with a NaCl gradient in the same buffer. The main peak of PRL from this column is monomeric, assessed by dynamic light scattering and chromatography on Superdex (22). The PRL solution was quick-frozen in a dry ice-ethanol bath in aliquots, and under these conditions it remains monomeric when thawed (22). The PRL concentration was determined by absorption at 280, using 0.1% = 0.9 OD units (34), and was confirmed by amino acid analysis at the W. M. Keck Foundation Biotechnology Resource Laboratory (New Haven, CT).

Assay for aggregation in solution
Aliquots of PRL were dialyzed against 0.001 M Tris/0.025 M KCl, pH 7.5, and then NaN₃ was added to a final concentration of 1 mM. Portions of this solution were diluted into a final volume of 50 µl containing 5 µM PRL, unless otherwise indicated,120 mM KCl, and 0.01 M HEPES, pH 7.4, unless otherwise indicated, or MES, pH 6.0 or 5.5, plus divalent cations and polyethylene glycol (Aldrich Chemical Co., Inc., Milwaukee, WI; average mol wt, 8000) as indicated. The solutions were incubated at 35 C for 15 min and centrifuged at 29,600 x g for 30 min at 4 C, and 30 µl of the supernatant were assayed for protein using Micro BCA from Pierce Chemical Co. (Rockford, IL). The data for soluble PRL are presented as a percentage of the total PRL added to the assay. In most of the assays for PRL aggregation in cells, we used more extensive centrifugation procedures, as performed in the original investigations of Lubrol insolubility (5), but we found that Lubrol-insoluble aggregates formed in cells would sediment under the conditions we used for this assay in solution (unpublished results).

Assay for aggregation in cells
We obtained adherent AtT20 cells from Dr. Sharon Milgram. AtT20 or COS cells, at 2.3 x 10⁵ cells/60-mm plate, were transfected with 15 µl Superfect (QIAGEN, Valencia, CA) and 5 µg pcDNA3 containing sequences for human wild-type PRL or H27A-PRL (35). For incorporation of ³⁵S-labeled amino acids, cells were incubated 1 d after transfection with 180 µC Express ³⁵S-Protein Labeling Mix (NEN Life Science Products, Boston, MA) in cysteine- and methionine-free DMEM with 10 mM MES, 10 mM HEPES, 4 mM NaHCO₃, and 5% horse serum (Central Biomedia, Inc., Irwin, MO). During the chase period cells were incubated with DMEM plus 2 mM cysteine, 2 mM methionine, 10 mM HEPES, 10 mM MES, 4 mM NaHCO₃, and 15% horse serum. Solutions used for these incubations were equilibrated in a 5% CO₂ atmosphere before use. For analysis at each time point, medium and cells were collected, and cells were dissolved in 0.32 M sucrose, 1.5% Lubrol, 0.5% BSA, 10 mM sodium phosphate buffer (pH 6.5), 100 µg/ml phenylmethylsulfonylfluoride, 1 µg/ml aprotinin, and 2.5 µg/ml pepstatin, then centrifuged at 50,000 x g for 1 h at 4 C. The pellet was resuspended in lysis buffer, and immunoprecipitation followed by gel electrophoresis was carried out as previously described (5). The amounts of [³⁵S]protein in the supernatant, pellet, and medium were quantified after gel electrophoresis using a Molecular Imager (Bio-Rad Laboratories, Inc., Hercules, CA). Antiserum to human PRL was obtained from the NIDDK National Hormone and Pituitary Program and A. F. Parlow.

Atomic absorption
Precipitates of human PRL were generated with the conditions detailed in Table 1. The precipitates were washed once with cold buffer solution without Zn²⁺ or Cu²⁺ and digested in 40 µl 1 M nitric acid at 54 C for 2 h. After dilution, appropriate cation concentrations were measured by flame atomic absorption spectroscopy using an Analyst 300 (Perkin-Elmer Corp., Norwalk, CT) equipped with a hollow cathode lamp (Perkin-Elmer Corp.) appropriate for each cation. Measurements were obtained with a 0.7-nm slit width, 15-mA lamp current, read delay of 0 sec, read time of 5 sec, and wavelengths of 213.9, 324.8, and 285.2 nm for zinc, copper, and magnesium, respectively. Calibration standards of 25, 50, 100, and 300 µg/dl were used for zinc and copper, and calibration standards of 10, 20, 60, and 100 µg/dl were used for magnesium. All precipitation conditions were performed on duplicate samples, with averages reported in Table 1. The amount of Zn²⁺ or Cu²⁺ precipitated in the absence of PRL was negligible. To assess the possibility of nonspecifically trapped cations being retained by the precipitates, 100 µM magnesium (which does not precipitate PRL alone) was added to some samples along with Zn²⁺ or Cu²⁺. Although copper and zinc were present in these samples in consistent amounts, no magnesium was detected.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

View larger version (24K): [in this window] [in a new window]   Figure 1. Solubility and secretion of human wild-type PRL or H27A-PRL in transiently transfected AtT20 cells. A, Incorporation of 35S-labeled amino acids into wild-type PRL or H27A-PRL after a 10-min pulse or after a 10-min pulse followed by a 30- or 60-min chase period in AtT20 cells transfected with wild-type PRL or H27A-PRL. S, Supernatant of a 50,000 x g centrifugation at 4 C of cells lysed in 1.5% Lubrol; P, pellet from the same centrifugation. M, Medium from the same sample. B, Top, [35S]PRL secreted into the medium after the 10-min pulse, as a percentage of the total [35S]PRL (intracellular plus extracellular) in the culture. Middle, Intracellular [35S]PRL that was soluble in Lubrol after the 10-min pulse as a percentage of the total [35S]PRL. Bottom, Intracellular [35S]PRL that was insoluble in Lubrol after the 10-min pulse as a percentage of total [35S]PRL in the culture. Data in B represent the mean of two or more experiments, and the bars represent the ranges or the SE. When no bars are shown, they fell within the symbols. , Wild-type PRL; , H27A-PRL.

View larger version (24K): [in this window] [in a new window]   Figure 2. Solubility of wild-type PRL in transiently transfected AtT20 and COS cells. A, Incorporation of 35S-labeled amino acids into wild-type PRL after a 10-min pulse or after a 10-min pulse followed by a 30- or 60-min chase period in COS cells transfected with wild-type PRL. S, Supernatant of a 50,000 x g centrifugation at 4 C of cells lysed in 1.5% Lubrol; P, pellet from the same centrifugation; M, medium from the same sample. Details are given in Fig. 1. B, Left panels, AtT20 cells were incubated with or without 30 µM chloroquine during the chase period, and [35S]PRL secreted into the medium (top panel) or remaining insoluble in the pellet (bottom panel) is shown as a percentage of the total [35S]PRL. , No chloroquine; , 30 µM chloroquine. B, Right panels, [35S]PRL in COS cells. Top panel, Amount secreted into the medium. Bottom panel, Amount in Lubrol-insoluble pellet.

Because the presence of Cu²⁺ and Zn²⁺ distinguishes the secretory pathway of neuroendocrine cells from those of many other cell types, we tested the aggregation of human PRL in AtT20 cells in the presence of diethyldithiocarbamate (DDC), a membrane-permeable compound that chelates Cu²⁺ and Zn²⁺ (40, 41). In three experiments, an average of 37 ± 1.6% of [³⁵S]PRL was aggregated by 15 min in cells without chelator, compared with 27 ± 1.1% (P < 0.05) in cells treated with DDC (Fig. 3). There was no effect on the secretion of PRL in cells treated with DDC (not shown).

View larger version (24K): [in this window] [in a new window]   Figure 3. [35S]PRL that is insoluble in transiently transfected AtT20 cells in the presence or absence of DDC. Cells were incubated with or without DDC. The graph summarizes three experiments with different conditions. Open symbols, No drug; filled symbols, DDC present before and during the pulse chase procedure; squares, Incubation with 2 µM DDC for 72 h before the pulse; circles, incubation with 2 µM DDC for 8.5 h before the pulse; triangles, incubation with 4 µM DCC for 4 h before the pulse.

View larger version (22K): [in this window] [in a new window]   Figure 4. Effects of pH, Ca2+, and PEG on human PRL solubility. PRL was incubated for 15 min at 35 C and centrifuged, and protein remaining in the supernatant was assayed as described in Materials and Methods. Each point is the mean of duplicate samples, and the error bars give the range. When no bars are shown, they are within the symbol. A, Effects of Ca2+ and PEG on PRL solubility. , No additions; , 90 mg/ml PEG; , 10 mM Ca2+; , 10 mM Ca2+ and 90 mg/ml PEG. B, Effect of PEG on PRL solubility. , pH 7.4; , pH 7.4, 1 mM EDTA; , pH 6.0; , pH 6.0, 1 mM EDTA; , pH 5.5; , pH 5.5, 1 mM EDTA.

Zn²⁺ and Cu²⁺ at physiologically relevant concentrations (42) caused PRL in solution to aggregate (Fig. 5). The exact concentrations required depended on the pH, and the pH dependence was different for Zn²⁺ and Cu²⁺ (Fig. 5). Cu²⁺ was most effective at pH 6, at which over 60% of PRL was precipitated by 20 µM (Fig. 5). Such concentrations of Cu²⁺ reduced PRL solubility at pH 7.4 in HEPES buffer (Fig. 5), but not in 3-(N-morpholino)propanesulfonic acid (MOPS) buffer (not shown), in which at least 5-fold more Cu²⁺ was required to obtain the same percent precipitation.

View larger version (23K): [in this window] [in a new window]   Figure 5. Effects of Zn2+ and Cu2+ concentrations on solubility of human wild-type PRL at pH 7.4, 6, and 5.5 in the presence or absence of PEG. Details are described in Fig. 4. A, Solubility in the presence of Zn2+. B, Solubility in the presence of Cu2+. Open symbols, In the absence of PEG; closed symbols, in the presence of PEG; circles, pH 7.4; squares, pH 6; triangles, pH 5.5. Left panel, Samples with added Zn2+; right panel, samples with added Cu2+.

Mutants of human GH that do not bind Zn²⁺ well do not self-associate well in the presence of Zn²⁺ (24). Human H27A-PRL, however, which also does not bind Zn²⁺ well (22), precipitated almost as well as wild-type PRL with Zn²⁺ in the presence and absence of PEG at pH 5.5 (Fig. 6) as well as 6.0 and 7.5, and in the presence of Cu²⁺ in all conditions (not shown). The ability of H27A-PRL to precipitate in the presence of Zn²⁺ shows that low affinity Zn²⁺-binding sites are sufficient for aggregation.

View larger version (23K): [in this window] [in a new window]   Figure 6. Effect of Zn2+ on solubility of wild-type and H27A-PRL in the presence and absence of PEG at pH 5.5. Open symbols, H27A-PRL; filled symbols, wild-type PRL; circles, in the absence of PEG; squares, in the presence of 90 mg/ml PEG.

View larger version (21K): [in this window] [in a new window]   Figure 7. Effects of Zn2+ and Cu2+ combinations on the solubility of wild-type PRL at pH 6.0 in the absence (top panel) or presence (bottom panel) of 90 mg/ml PEG. The micromolar concentrations of Zn2+ and Cu2+ are indicated in the labels below the bars.

View larger version (16K): [in this window] [in a new window]   Figure 8. Reversibility of precipitation of wild-type PRL by Zn2+, Cu2+, or PEG. PRL solutions were incubated with 80 µM Zn2+, pH 7.4 (left panel); 50 µM Cu2+, pH 7.4 (middle panel); or 180 mg/ml PEG, pH 6.0 (right panel), and centrifuged after the incubation, and the amount of PRL remaining in the supernatant was determined (leftmost bar in each panel). The pellets, after the supernatants were removed, were incubated for 10 min in 120 mM KCl and 10 mM HEPES, pH 7.4, with 0, 0.1, or 1 mM EDTA as indicated under the bars. This solution was then recentrifuged, and the supernatant was assayed for PRL that had become soluble.

Cu²⁺ or Zn²⁺ cause the amyloid ß-peptide to aggregate by forming intermolecular cross-bridges (43). If these divalent cations cause large aggregates to form by making connections with more than one PRL molecule so that a large interconnected network forms, then increasing PRL concentrations should be able to prevent precipitation because excess PRL should reduce the probability of cross-bridging. In the absence of PEG, increasing the amount of PRL prevented precipitation caused by a constant amount of cation, either 25 µM Cu²⁺ at pH 6 or 20 µM Zn²⁺ at pH 7.4 (Fig. 9A). More than one Cu²⁺ and Zn²⁺ was precipitated for each PRL molecule under these conditions (Table 1). These data are consistent with Cu²⁺ and Zn²⁺ forming PRL aggregates by cross-bridging. In 90 mg/ml PEG at pH 7.4, the results are similar, and increasing amounts of PRL prevented precipitation (Fig. 4B). In 90 mg/ml PEG at pH 6, however, inhibition by increasing PRL was attenuated, and adding more PRL did not abolish precipitation caused by 10 µM Zn²⁺ (Fig. 9C). Under these conditions 10 µM Zn²⁺ formed a precipitate with PRL that had 0.5 µM Zn²⁺/PRL molecule (Table 1). In 90 mg/ml PEG at pH 6, therefore, the mechanism of precipitation by Zn²⁺ is more complex than simple cross-bridging. The difference is the presence of PEG as well as the acidic pH, and not just the acidic pH alone, for although increasing PRL did not prevent precipitation caused by 80 µM Zn²⁺ at pH 6 in the absence of PEG (Fig. 9A), higher Zn²⁺ concentrations were necessary to precipitate PRL at this pH than at pH 7.4, and the ratio of Zn²⁺ to PRL was still greater than 1 under these conditions (Table 1).

View larger version (22K): [in this window] [in a new window]   Figure 9. Effects of increasing PRL concentrations on solubility of wild-type PRL in the presence of Zn2+ or Cu2+ in the presence or absence of PEG. A, Effect in the absence of PEG. , 20 µM Zn2+, pH 7.4; , 80 µM Zn2+, pH 6.0; , 25 µM Cu2+, pH 6.0. B, Effect at pH 7.4 in the presence of PEG. Triangle, 90 mg/ml PEG, pH 7.4; squares, 10 µM Zn2+ and 90 mg/ml PEG, pH 7.4; circles, 15 µM Cu2+ and 90 mg/ml PEG, pH 7.4. C, Effect at pH 6.0 in the presence of PEG. Triangles, 90 mg/ml PEG, pH 6.0; squares, 10 µM Zn2+ and 90 mg/ml PEG, pH 6.0; circles, 7 µM Cu2+ and 90 mg/ml PEG, pH 6.0.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
The results presented here are consistent with the following model for human PRL aggregation in neuroendocrine cells. Human PRL is soluble immediately after synthesis in the endoplasmic reticulum. Concentrations of PRL are higher in the Golgi complex than in the endoplasmic reticulum (44), and the pH decreases to close to 6 in the trans-Golgi lumen, which causes PRL to self-associate. Intermolecular cross-bridging of PRL oligomers through low affinity bonds with Zn²⁺ and Cu²⁺, which are also present in the trans-Golgi lumen, enhance aggregation at acidic pH. The results in neuroendocrine cells that support this model are 1) the pH dependency of human PRL, demonstrated by the sensitivity to chloroquine, 2) the previously published evidence that Cu²⁺ and Zn²⁺ are in the secretory pathway of neuroendocrine cells, 3) the ability of the chelator DDC to slow aggregation, and 4) the similarity of aggregation of human wild-type PRL and H27A-PRL, indicating that high affinity (1 µM) Zn²⁺ binding is not necessary. The effect of the chelator DDC is small, but may underestimate the roles of Zn²⁺ and Cu²⁺. It is likely to be difficult to remove Zn²⁺ and Cu²⁺ entirely from the secretory pathway without causing cellular toxicity, because these divalent cations are necessary for so many cellular functions, and we were unable to use higher concentrations of DDC without problems with toxicity.

The dependence of human PRL aggregation in cells on a mildly acidic pH differs from that of rat PRL. Rat PRL aggregations slows, but still occurs in the presence of 30 µM chloroquine (5). Species specificity is also found in the solubility of rat and human PRL in solution. Rat PRL precipitates when incubated at neutral pH (17), an effect we did not find with human PRL even in the presence of high concentrations of PEG. The behavior of rat PRL in solution is consistent with its behavior in cells, where aggregation occurs without acidic pH. Rat and human PRL have similar structures, but only have 122 identical amino acids of 199 (human PRL) (45). Both are acidic proteins, as are most secretory granule proteins. Until more is understood about why such acidic proteins self-associate, it is not possible to propose rationally which differences between rat and human PRL are important.

We found that Cu²⁺ and Zn²⁺ precipitate PRL in solution. Cu²⁺ or Zn²⁺ do not precipitate all secretory proteins as easily as human PRL; serum albumin remains soluble in the presence of Cu²⁺ or Zn²⁺ (46, 47) under conditions similar to those used in these investigations, and we found no effect of Cu²⁺ or Zn²⁺ on albumin solubility in the presence of PEG (not shown). Other divalent cations do not have effects at physiologically relevant concentrations. The binding of Zn²⁺ and Cu²⁺ therefore is interesting because there is some specificity, even though the sites are of relatively low affinity.

Concentrations of Zn²⁺ found in serum precipitate human PRL in solution at pH 7.4, but become less effective at lower pH, a result that did not suggest a strong role for Zn²⁺ in causing aggregation in cells. In the presence of PEG, however, the ability of Zn²⁺ to form precipitates at acidic pH is greatly enhanced. Zn²⁺ binds to several amino acids, including glutamate, aspartate, glutamine, and histidine (43). The decrease in PRL precipitation by Zn²⁺ as the pH becomes acidic in the absence of PEG is consistent with an important involvement of Zn²⁺:histidine binding, because protonation of histidine prevents Zn²⁺ binding to it (43). The simplest explanation for the difference in the ability of Zn²⁺ to precipitate PRL at acidic pH in the presence and absence of PEG is that Zn²⁺ forms complexes at pH 6 in the presence of PEG primarily with amino acids other than histidine, such as glutamate and aspartate (43), and in the absence of PEG primarily with histidine. The ability to form cross-bridges with other amino acids may be enhanced by the presence of PEG more than the ability to form them with histidine, so that in the presence of PEG, histidine binding no longer predominates in forming aggregates.

Aggregation of human PRL in cells and in solution does not require the Zn²⁺-binding site that has been characterized (22), as H27A-PRL behaves very much as the wild type in all the conditions we examined. Relatively low affinity Zn²⁺ binding is sufficient to cause precipitation in solution. In this respect PRL differs from human GH, in which disruption of the Zn²⁺-binding site greatly reduced self-association in solution (24). Whether the Zn²⁺-binding site is necessary for GH aggregation in cells has not been determined.

Cu²⁺ forms a complex with histidine with a higher affinity than H⁺ (43), so mildly decreasing the pH does not affect Cu²⁺ binding as it does that of Zn²⁺, and the ability of Cu²⁺ to precipitate PRL well at pH 6 and 5.5 does not rule out a role for histidine residues in Cu²⁺ binding. The sensitivity to Cu²⁺ depends on the buffer used. MOPS has no reported action with Cu²⁺ (48), but several investigations have shown interactions of HEPES and Cu²⁺ (48, 49, 50), and formation of a complex of Cu²⁺ and HEPES may keep Cu²⁺ in a soluble form (50). Redox-active transition metal cations, such as Cu²⁺, cause oxidative reactions. Cu²⁺ is usually in a complex in yeast cells; there is less than one free Cu²⁺ ion in an entire yeast cell (51), which helps protect against unwanted oxidative reactions. Although the form in which Cu²⁺ exists in the secretory pathway is unknown, it may also exist there in part in weakly bound complexes. The complex of Cu²⁺ bound weakly to HEPES may resemble the state of Cu²⁺ in the secretory pathway more closely than free Cu²⁺ alone.

The means by which crowding by addition of agents such as PEG affects reaction rates and equilibria have been investigated for a number of reactions, and in all cases to date the effects can be related to the volume available to large components of the reaction (29, 30, 31). Available volume is defined as the volume that may be occupied by the center of mass of a molecule. The effect of crowding on the available volume of a relatively large species, such as a protein, is much greater than the effect on a relatively small species, such as metal ions, and so the available volume for PRL cannot be calculated simply by subtracting the volume occupied by the crowding agent (29, 30, 31). From our investigations, we cannot say whether the effect on PRL is only caused by changing the available volume; to determine this would require much more extensive investigations of the thermodynamic activities of PRL self-association in the presence and absence of PEG. Since reducing the available volume has accounted for the increase in activities in other protein association reactions, it seems plausible that it accounts for the effects on PRL precipitation as well.

Regardless of the mechanism, the precipitates formed in the presence of PEG contain less than one divalent cation per PRL molecule. This ratio is important because aggregates of PRL in cells are unlikely to contain more than that. Cunningham et al. (24) have estimated that the human pituitary gland contains enough Zn²⁺ to form a 1:1 complex of GH and Zn²⁺ if all the Zn²⁺ in the gland is in the secretory granules of somatotrophs; the human male pituitary gland contains an average of 340 nmol GH (52) and 340 nmol Zn²⁺ (53), assuming that the gland is 1 g and 65% water (24). Because all Zn²⁺ is not in somatotrophs in the pituitary gland, and all Zn²⁺ is not in granules in somatotrophs, even a 1:1 complex is not feasible. There is less Cu²⁺, an average of 66 nmol in a human male pituitary gland (54). In men, there is 50 times more GH than PRL (52), so there might be enough of these ions to complex PRL in men, but in women, serum PRL levels increase over 50-fold during pregnancy, and presumably pituitary PRL content also increases greatly at the same time. It is therefore unlikely that intermolecular bridging by Cu²⁺ or Zn²⁺ is solely responsible for the aggregation of PRL and GH in pituitary cells. In the presence of PEG at pH 6.0, PRL precipitates must form by another means, because, with less than one Zn²⁺ per PRL molecule, there cannot be cross-bridges connecting many molecules. In these conditions, Zn²⁺ may interact with PRL to have a seeding effect or Zn²⁺ may cross-bridge PRL oligomers that self-associate in the presence of PEG at pH 6. Either mechanism could enhance PRL aggregation in cells.

Precipitates of PRL in the absence of PEG do not therefore appear to resemble aggregates in cells, because there is no precipitation with acidic pH alone, precipitation with Zn²⁺ is most effective at neutral pH, and the ratios of divalent cations to PRL are too high. The precipitates in the presence of PEG show pH dependency alone and in the presence of Zn²⁺ that are more consistent with pH dependency demonstrated using chloroquine in intact cells, and the ratios of divalent cations to PRL are more appropriate. The aggregates obtained in PEG are therefore better models for investigations of structural characteristics than those obtained in its absence. The failures of some investigations of aggregation in solution to be consistent with storage in secretory granules (55) may be because the conditions used for aggregation in solution did not mimic the kinds of aggregates that occur in cells. Recognition of the properties of hormone aggregate by secretory granule membrane proteins may be an important step in forming secretory granules (55). Obtaining the appropriate aggregates in solution is an important step to characterize the properties that may be recognized.

	Acknowledgments

 
We thank Florecita Santos for help with the atomic absorption.

	Footnotes

 
This work was supported by NIH Grant DK-46807 and a grant from the American Diabetes Association.

Abbreviations: DDC, Diethyldithiocarbamate; PEG, polyethylene glycol.

Received October 22, 2001.

Accepted for publication December 14, 2001.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Giannattasio G, Zanini A, Meldolesi J 1975 Molecular organization of rat prolactin granules. I. In vitro stability of intact and "membraneless" granules. J Cell Biol 64:246–251[Abstract/Free Full Text]
2. Michael J, Carroll R, Swift HH, Steiner DF 1987 Studies on the molecular organization of rat insulin secretory granules. J Biol Chem 262:16531–16535[Abstract/Free Full Text]
3. Rambourg A, Clermont Y, Chretien M, Olivier L 1992 Formation of secretory granules in the Golgi apparatus of prolactin cells in the rat pituitary gland: a stereoscopic study. Anat Rec 232:169–179[CrossRef][Medline]
4. Mastro RM, Dannies PS 1995 Lack of correlation of distribution of prolactin (PRL) charge isoforms with induction of PRL storage. Endocrinology 136:69–74[Abstract]
5. Lee MS, Zhu YL, Chang JE, Dannies PS 2001 Acquisition of Lubrol insolubility, a common step for growth hormone and prolactin in the secretory pathway of neuroendocrine cells. J Biol Chem 276:715–721[Abstract/Free Full Text]
6. Anderson RG, Pathak RK 1985 Vesicles and cisternae in the trans Golgi apparatus of human fibroblasts are acidic compartments. Cell 40:635–643[CrossRef][Medline]
7. Orci L, Ravazzola M, Amherdt M, Madsen O, Perrelet A, Vassalli JD, Anderson RG 1986 Conversion of proinsulin to insulin occurs coordinately with acidification of maturing secretory vesicles. J Cell Biol 103:2273–2281[Abstract/Free Full Text]
8. Seksek O, Biwersi J, Verkman AS 1995 Direct measurement of trans-Golgi pH in living cells and regulation by second messengers. J Biol Chem 270:4967–4970[Abstract/Free Full Text]
9. Demaurex N, Furuya W, D’Souza S, Bonifacino JS, Grinstein S 1998 Mechanism of acidification of the trans-Golgi network (TGN). In situ measurements of pH using retrieval of TGN38 and furin from the cell surface. J Biol Chem 273:2044–2051[Abstract/Free Full Text]
10. Wu MM, Grabe M, Adams S, Tsien RY, Moore HP, Machen TE 2001 Mechanisms of pH regulation in the regulated secretory pathway. J Biol Chem 276:33027–33035[Abstract/Free Full Text]
11. Rindler MJ 1998 Carboxypeptidase E, a peripheral membrane protein implicated in the targeting of hormones to secretory granules, co-aggregates with granule content proteins at acidic pH. J Biol Chem 273:31180–31185[Abstract/Free Full Text]
12. Song L, Fricker LD 1995 Calcium and pH-dependent aggregation of carboxypeptidase E. J Biol Chem 270:7963–7967[Abstract/Free Full Text]
13. Colomer V, Kicska GA, Rindler MJ 1996 Secretory granule content proteins and the luminal domains of granule membrane proteins aggregate in vitro at mildly acidic pH. J Biol Chem 271:48–55[Abstract/Free Full Text]
14. Chanat E, Huttner WB 1991 Milieu-induced, selective aggregation of regulated secretory proteins in the trans-Golgi network. J Cell Biol 115:1505–1519[Abstract/Free Full Text]
15. Freedman SD, Scheele GA 1993 Regulated secretory proteins in the exocrine pancreas aggregate under conditions that mimic the trans-Golgi network. Biochem Biophys Res Commun 197:992–999[CrossRef][Medline]
16. Leblond FA, Viau G, Laine J, Lebel D 1993 Reconstitution in vitro of the pH-dependent aggregation of pancreatic zymogens en route to the secretory granule: implication of GP-2. Biochem J 291:289–296
17. Thompson ME, Zimmer WE, Haynes AL, Valentine DL, Forss-Petter S, Scammell JG 1992 Prolactin granulogenesis is associated with increased secretogranin expression and aggregation in the Golgi apparatus of GH₄C₁ cells. Endocrinology 131:318–326[Abstract]
18. Canaff L, Brechler V, Reudelhuber TL, Thibault G 1996 Secretory granule targeting of atrial natriuretic peptide correlates with its calcium-mediated aggregation. Proc Natl Acad Sci USA 93:9483–9487[Abstract/Free Full Text]
19. Pezzati R, Bossi M, Podini P, Meldolesi J, Grohovaz F 1997 High-resolution calcium mapping of the endoplasmic reticulum-Golgi-exocytic membrane system. Electron energy loss imaging analysis of quick frozen-freeze dried PC12 cells. Mol Biol Cell 8:1501–1512[Abstract]
20. Gorr SU, Shioi J, Cohn DV 1989 Interaction of calcium with porcine adrenal chromogranin A (secretory protein-I) and chromogranin B (secretogranin I). Am J Physiol 257:E247–E254
21. Thorlacius-Ussing O 1987 Zinc in the anterior pituitary of rat: a histochemical and analytical work. Neuroendocrinology 45:233–242[CrossRef][Medline]
22. Sun Z, Li PS, Dannies PS, Lee JC 1996 Properties of human prolactin (PRL) and H27A-PRL, a mutant that does not bind Zn⁺⁺. Mol Endocrinol 10:265–271[Abstract]
23. Permyakov EA, Veprintsev DB, Deikus GY, Permyakov SE, Kalinichenko LP, Grishchenko VM, Brooks CL 1997 pH-induced transition and Zn²⁺-binding properties of bovine prolactin. FEBS Lett 405:273–276[CrossRef][Medline]
24. Cunningham BC, Mulkerrin MG, Wells JA 1991 Dimerization of human growth hormone by zinc. Science 253:545–548[Abstract/Free Full Text]
25. Glembotski CC, Eipper BA, Mains RE 1984 Characterization of a peptide -amidation activity from rat anterior pituitary. J Biol Chem 259:6385–6392[Abstract/Free Full Text]
26. Watanabe T, Jeziorowski T, Wuttke W, Grube D 1993 Secretory granules and granins in hyperstimulated male rat gonadotropes. J Histochem Cytochem 41:1801–1812[Abstract]
27. Kako K, Tsumori K, Ohmasa Y, Takahashi Y, Munekata E 2000 The expression of cox17p in rodent tissues and cells. Eur J Biochem 267:6699–6707[Medline]
28. Hutton JC, Penn EJ, Peshavaria M 1983 Low-molecular-weight constituents of isolated insulin-secretory granules. Bivalent cations, adenine nucleotides and inorganic phosphate. Biochem J 210:297–305[Medline]
29. Minton AP 2001 The influence of macromolecular crowding and macromolecular confinement on biochemical reactions in physiological media. J Biol Chem 276:10577–10580[Free Full Text]
30. Ellis RJ 2001 Macromolecular crowding: an important but neglected aspect of the intracellular environment. Curr Opin Struct Biol 11:114–119[CrossRef][Medline]
31. Minton AP 2000 Implications of macromolecular crowding for protein assembly. Curr Opin Struct Biol 10:34–39[CrossRef][Medline]
32. Paris N, Rentier-Delrue F, Defontaine A, Goffin V, Lebrun JJ, Mercier L, Martial JA 1990 Bacterial production and purification of recombinant human prolactin. Biotechnol Appl Biochem 12:436–449[Medline]
33. Rhee HK, Sun Z, Kim SS, Goffin V, Martial JA, Dannies PS 1995 Biological activity and immunological reactivity of human prolactin mutants. Endocrinology 136:4990–4995[Abstract]
34. Cunningham BC, Henner DJ, Wells JA 1990 Engineering human prolactin to bind to the human growth hormone receptor. Science 247:1461–1465[Abstract/Free Full Text]
35. Sun Z, Lee MS, Rhee HK, Arrandale JM, Dannies PS 1997 Inefficient secretion of human H27A-prolactin, a mutant that does not bind Zn²⁺. Mol Endocrinol 11:1544–1551[Abstract/Free Full Text]
36. Mains RE, May V 1988 The role of a low pH intracellular compartment in the processing, storage, and secretion of ACTH and endorphin. J Biol Chem 263:7887–7894[Abstract/Free Full Text]
37. Hiebsch RR, Raub TJ, Wattenberg BW 1991 Primaquine blocks transport by inhibiting the formation of functional transport vesicles. Studies in a cell-free assay of protein transport through the Golgi apparatus. J Biol Chem 266:20323–20328[Abstract/Free Full Text]
38. Braulke T, Geuze HJ, Slot JW, Hasilik A, von Figura K 1987 On the effects of weak bases and monensin on sorting and processing of lysosomal enzymes in human cells. Eur J Cell Biol 43:316–321[Medline]
39. Stoller TJ, Shields D 1989 The role of paired basic amino acids in mediating proteolytic cleavage of prosomatostatin. Analysis using site-directed mutagenesis. J Biol Chem 264:6922–6928[Abstract/Free Full Text]
40. Mains RE, Park LP, Eipper BA 1986 Inhibition of peptide amidation by disulfiram and diethyldithiocarbamate. J Biol Chem 261:11938–11941[Abstract/Free Full Text]
41. Morpurgo L, Desideri A, Rigo A, Viglino P, Rotilio G 1983 Reaction of N,N-diethyldithiocarbamate and other bidentate ligands with Zn, Co and Cu bovine carbonic anhydrases. Inhibition of the enzyme activity and evidence for stable ternary enzyme-metal-ligand complexes. Biochim Biophys Acta 746:168–175[CrossRef][Medline]
42. Iyengar GV, Kollmer WE, Bowen HJM 1978 The elemental composition of human tissues and body fluids. New York: Verlag Chemie
43. Miura T, Suzuki K, Kohata N, Takeuchi H 2000 Metal binding modes of Alzheimer’s amyloid ß-peptide in insoluble aggregates and soluble complexes. Biochemistry 39:7024–7031[CrossRef][Medline]
44. Farquhar MG, Reid JJ, Daniell LW 1978 Intracellular transport and packaging of prolactin: a quantitative electron microscope autoradiographic study of mammotrophs dissociated from rat pituitaries. Endocrinology 102:296–311[Abstract]
45. Nicoll CS, Mayer GL, Russell SM 1986 Structural features of prolactins and growth hormones that can be related to their biological properties [published erratum appears in Endocr Rev 1987 Feb;8(1):43]. Endocr Rev 7:169–203[Medline]
46. Atwood CS, Moir RD, Huang X, Scarpa RC, Bacarra NM, Romano DM, Hartshorn MA, Tanzi RE, Bush AI 1998 Dramatic aggregation of Alzheimer ß by Cu(II) is induced by conditions representing physiological acidosis. J Biol Chem 273:12817–12826[Abstract/Free Full Text]
47. Goumakos W, Laussac JP, Sarkar B 1991 Binding of cadmium(II) and zinc(II) to human and dog serum albumins. An equilibrium dialysis and 113Cd-NMR study. Biochem Cell Biol 69:809–820[Medline]
48. Yu Q, Kandegedara A, Xu Y, Rorabacher DB 1997 Avoiding interferences from Good’s buffers: a contiguous series of noncomplexing tertiary amine buffers covering the entire range of pH 3-11. Anal Biochem 253:50–56[CrossRef][Medline]
49. Vasconcelos MT, Azenha MA, Lage OM 1996 Electrochemical evidence of surfactant activity of the hepes pH buffer which may have implications on trace metal availability to cultures in vitro. Anal Biochem 241:248–253[CrossRef][Medline]
50. Simpson JA, Cheeseman KH, Smith SE, Dean RT 1988 Free-radical generation by copper ions and hydrogen peroxide. Stimulation by Hepes buffer. Biochem J 254:519–523[Medline]
51. Rae TD, Schmidt PJ, Pufahl RA, Culotta VC, O’Halloran TV 1999 Undetectable intracellular free copper: the requirement of a copper chaperone for superoxide dismutase. Science 284:805–808[Abstract/Free Full Text]
52. Thorner MO, Vance ML, Horvath E, Kovacs K 1992 Textbook of endocrinology, 8th ed. Philadelphia: Saunders
53. Thorlacius-Ussing O, Gregersen M, Hertel N 1988 The concentration of twelve elements in the anterior pituitary from human subjects and rats as measured by particle induced X-ray emission (PIXE). Biol Trace Elem Res 16:189–202[Medline]
54. Arias AE, Velez-Granell CS, Mayer G, Bendayan M 2000 Colocalization of chaperone cpn60, proinsulin and convertase PC1 within immature secretory granules of insulin-secreting cells suggests a role for cpn60 in insulin processing. J Cell Sci 113:2075–2083[Abstract]
55. Dannies PS 1999 Protein hormone storage in secretory granules: mechanisms for concentration and sorting. Endocr Rev 20:3–21[Abstract/Free Full Text]

This article has been cited by other articles:

				D. I Iliev, N. E Wittekindt, M. B Ranke, and G. Binder In vitro analysis of hGH secretion in the presence of mutations of amino acids involved in zinc binding J. Mol. Endocrinol., August 1, 2007; 39(2): 163 - 167. [Abstract] [Full Text] [PDF]

				Y. L. Zhu, A. Abdo, J. F. Gesmonde, K. C. Zawalich, W. Zawalich, and P. S. Dannies Aggregation and Lack of Secretion of Most Newly Synthesized Proinsulin in Non-{beta}-Cell Lines Endocrinology, August 1, 2004; 145(8): 3840 - 3849. [Abstract] [Full Text] [PDF]

				L. McGuinness, C. Magoulas, A. K. Sesay, K. Mathers, D. Carmignac, J.-B. Manneville, H. Christian, J. A. Phillips III, and I. C. A. F. Robinson Autosomal Dominant Growth Hormone Deficiency Disrupts Secretory Vesicles in Vitro and in Vivo in Transgenic Mice Endocrinology, February 1, 2003; 144(2): 720 - 731. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sankoorikal, B.-J. Articles by Dannies, P. S. Search for Related Content PubMed